CCA's outline of adding new patrons before the semester:
pipenv install && pipenv shellOn a regular basis, we sync names from Workday and card number changes from the TouchNet report to Koha, so that patrons who changed their preferred names or lost or changed their CCA ID cards don't have to update their account themselves.
pipenv run getdatapipenv run ./patron_update.py -p prox_report.csv -w data.json | tee -a prox_update.log where data.json is one of the (employee or student) Workday files.Before each semester, we load new patron accounts using data sourced from Workday to create a CSV that's then batch loaded into Koha.
Download JSON files from Google Cloud with pipenv run getdata. For a summer term, get pre-college students too with pipenv run getpcdata. Our scripts expect the JSON files to retain their names, e.g. "student_data.json". Download the report of "Prox" numbers (Custom Reports > "Accounts with Prox IDs").
Check that there are no new student majors not represented in "koha_mappings.py". The script "new-programs.sh" (requires jq) parses the employee/student data and writes all major/department values to text files in the data directory, then it runs git diff against its own prior iterations.
Run the main script python create_koha_csv.py prox_report.csv --end 2023-12-12 where the CSV is the prox report and the --end parameter is the last day of the semester (see Portal's Academic Calendar). Expiration dates for all account types (staff, student, faculty) are based on the end date. The script prints diagnostic messages for users with ambiguous accounts, often hourly or special programs instructors. We need to double check that these accounts either already exist or aren't needed.
On Koha's staff side, select Tools & then Import Patrons. Use the following settings:
After import, Koha informs you how many patrons were created & if any rows in the import CSV were malformed. You can copy the full text output of this page and save it into the data directory. You may need to check some duplicate card numbers; username changes not reflected in Koha is a common issue.
There's a pipenv run clean script to delete the data files, if they have their expected names, to run after the import is done.
Koha has a REST API with a /patrons endpoint. Read its documentation at https://library-staff.cca.edu/api/v1/.html
We could use a script similar to koha_patron/add_demo.py (WIP) to add patrons to Koha one-by-one with the API rather than in bulk with a CSV. Rather than use two scripts, patron_update.py could simply create missing patrons on the fly.
The API previously had a limitation that patron extended attributes could not be created nor modified. We use attributes to record student major and faculty department, so that curbed the API's usefulness. Luckily, a new /patron/{id}/extended_attributes route (see bug #23666) was added in Koha 21.05. We use the API in the "patron_update.py" script to update existing patron records without overwriting them entirely.
If we do the logical thing of GETting a patron record from the API, modifying it, then PUTting it back, Koha throws an error because there are read-only fields in the record. Remove them before sending the record back:
from koha_patron.patron import PATRON_READ_ONLY_FIELDS
for field in PATRON_READ_ONLY_FIELDS:
patron.pop(field) # patron = dict of the patron record